M9 noise performance

[jaapv]

Advertisement (gone after registration)

These days though, is it not so that most people view photographs on a screen? Of course, we can print digital files, but I think they are designed to be seen on a monitor. So it is important if noise shows up on monitor. If it will be different on a print, this is not such relevance for most people.

In that case why waste your money? An IPhone suffices.

[rosuna]

The problem with "noise" is not the grainy look of the images, or the color patches. Th problem with "noise" is the loss of detail, dynamic range, color gamut and tonal range. You cannot recover that. During post processing, you might eliminate or make a lift to the grainy texture (losing additional detail), but not improve the image.

 

There is something more in the last generation CMOS sensors from Canon and Nikon than just "post processing". The noise levels of the Leica CCD based cameras is higher, meaning lower dynamic range, tonal resolution, etc. at higher ISO values. The DxO charts show this very clearly. The image quality at lower ISO values is very good, but not better than that from those CMOS units. There are perceptible differences and you may prefer the unprocessed colors and sharpness or the noise (read: "grain") pattern of the CCD. That is all.

 

All these discussions about noise, about how good Lightroom is "suppressing" noise or how you can get Canon or Nikon image quality at high ISO by means of post processing is misleading, in my opinion. The facts are the facts.

[Advertisement]

The Leica Q3 is available for pre-order at B&H Photo Video and Adorama

[lct]

Well said Ruben.

[roydonian]

Heavy is impertinent. If you can't carry the camera, then pity you.

 

Passing years may change your view of the meaning of 'heavy'. My D200 and its lens weighed 1.7 kg. My M8.2 and TriElmar weigh 0.9 kg. My 68-year-old wrists certainly appreciate the lighter load.

 

Ten years ago I carried a Leicaflex + M5 + lenses in a bag that weighed 4.5 kg and thought nothing of it These days I try to keep the M8.2 + lenses & accessories to a load of less than 3 kg.

 

But I was always the sort of guy that they flicked sand at in the old Charles Atlas adverts.

 

Best regards,

 

Doug

[wlaidlaw]

Doug,

 

The M9 must be a bit heavier. My M9 with grip, Thumbie, Tri-Elmar and hood, comes up to 1027 grams. Obviously I am going to have to go on weight lifting classes ;-}} I do notice however that it seems quite heavy when I have the chrome/brass 35 ASPH Summilux on it.

 

Wilson

[kdriceman]

The problem with "noise" is not the grainy look of the images, or the color patches. Th problem with "noise" is the loss of detail, dynamic range, color gamut and tonal range. You cannot recover that. During post processing, you might eliminate or make a lift to the grainy texture (losing additional detail), but not improve the image.

 

There is something more in the last generation CMOS sensors from Canon and Nikon than just "post processing". The noise levels of the Leica CCD based cameras is higher, meaning lower dynamic range, tonal resolution, etc. at higher ISO values. The DxO charts show this very clearly. The image quality at lower ISO values is very good, but not better than that from those CMOS units. There are perceptible differences and you may prefer the unprocessed colors and sharpness or the noise (read: "grain") pattern of the CCD. That is all.

 

All these discussions about noise, about how good Lightroom is "suppressing" noise or how you can get Canon or Nikon image quality at high ISO by means of post processing is misleading, in my opinion. The facts are the facts.

 

All true enough, but it is still about tradeoffs. There are venues where I would opt for the DSLR because I can and it will be able to retain the detail at high ISO better. But for most of my high ISO shots I will still prefer the less obtrusive M9. Suggesting LR suppresses noise or is a magical fix is misleading to the extent people say that. I know LR doesn't "suppress" noise, but it does help finesse it and allows you to choose the extent to which you make the tradeoffs between loss of detail and grain.

[Jamie Roberts]

Advertisement (gone after registration)

{snipped}

 

All these discussions about noise, about how good Lightroom is "suppressing" noise or how you can get Canon or Nikon image quality at high ISO by means of post processing is misleading, in my opinion. The facts are the facts.

 

Has anyone said you get Nikon noise or DR at high ISOs? I find that misleading, actually.

 

You're right about dropping DR as you go up in ISO (DxO aside--they're doing something weird however in their tests). But you'd also be right on that with a Canon or Nikon--they also loose gradation, and tonality, and detail too.

 

The point is that my D3 "loses" it between one and two stops higher than my M9. Not 4 stops, not 3 or more. About 1.5... and that's significant yes.

 

The other point is colour shift: Nikons in particular are all over the place as the ISO increases. Yes, I can shoot at ISO 3200 and produce a nice print. I can push it to 6400 if I have to. But by then the files are mush and colour is terrible.

 

So I'll stand by what I said previously: Up to ISO 1600 on the M9 and I don't worry about it at all--it's better than the M8--no doubt about it. ISO 2500 on the M9 requires careful exposure and lack of DR mitigating choices, but if you've ever shot print film you're doing the same thing

 

Nikon territory ranges above that, so that even at equal ISOs (the M9 is 1/3 stop more light sensitive anyway) I have 1.5 more stops of exposure in the shadows if I want it... but generally I don't for a print if I haven't made a mistake in exposure.

 

No-one is saying they're equivalent technically. But practically, they are not as far apart as some people think and there are tradeoffs for really high ISO response. And those are facts too

Edited September 9, 2011 by Jamie Roberts

[sblitz]

i lived with 400asa film as the top end w/o too much "film noise" for seemingly forever, not sure why not being able to go above 1600 or 2500 or 25000 is such a major event. going to 1600 is amazing enough to me but rarely necessary, and i live in nyc not a place with constant bright sunshine (unless you are comparing to london ) . i pretty much try not to push the m9 past 800 as it is rarely needed so all of these ultra high iso arguments are technically interesting but they are a lot like arguing about how many angels can dance on the head of a pin.

[Jamie Roberts]

{snipped} i pretty much try not to push the m9 past 800 as it is rarely needed so all of these ultra high iso arguments are technically interesting but they are a lot like arguing about how many angels can dance on the head of a pin.

 

Except that not all of us are shooting outdoors... When you're in the business of shooting events indoors with very low available light, higher ISO response is a must. Even adding light in situations like that is tricky without an ISO capability higher than 400 or 800 because unless you have a lighting crew, it looks pretty awful.

 

Everyone's cutoff is different. If I could have an extra stop or a bit more on the M10 that's just about what I need for 99.99% of the things I shoot.

[wlaidlaw]

I do wonder, in its current dire circumstances, how much Kodak is prepared (or is allowed by its bankers) to spend on R&D on upgrading the M9's sensor. I personally would much prefer that Kodak targeted lower noise than higher pixels. I agree with Jamie, for indoor or evening event shots, where unless you are O. Winston Link, flash is impractical, even an extra EV of sensitivity would be very useful.

 

Given the experience that Kodak has in producing the very specialist DRF sensor, it is probably not a viable proposition for Leica to look elsewhere, unless they are prepared to delay the introduction of the M10 for a year or two. I would guess therefore, that we will only see modest improvements in noise on future M's, unless a wholly new sensor technology emerges, which is always possible.

 

Wilson

[hoppyman]

Andy I have no idea or comment on the other cameras. I'm not following you here on the comment regarding White Balance affecting noise though.

Assuming DNG , the value set in the camera is nothing more than an instructionrecorded with no effect on the actual data. It can be varied without penalty in post of course.

Unless you are actually referring to underexposure in one channel which you are needing to correct in post? Did you mean colour rendiring according to a preferred profile/output rather than white balance?

 

.....

Additionally, unless your test shot was under actual daylight, correcting for white balance from other types of light adds its own noise penalty.

 

The Leica/Kodak sensors have very strong blue filters in the Bayer pattern, which makes for very pure, saturated colors under sunlight (my Canon requires a big boost in saturation to match the Leica). But under yellowish indoor lights, the blue channel will be practically unexposed - and dragging it up to match the other colors will increase the chroma noise.

 

Again, a tradeoff. On the whole, I'll accept Leica's noise to get Leica's colors.

[sblitz]

Jamie -- I absolutely respect your view on this, considering this is your profession and you have been doing this for more than 10 minutes. My question is, in the days of film, what did you do? and can't m9 iso of 1600 suffice today?

[roydonian]

Doug,

 

The M9 must be a bit heavier. My M9 with grip, Thumbie, Tri-Elmar and hood, comes up to 1027 grams.

 

The M9 is about 50 grammes neavier than the M8, and your hood and grip would make their own additions to the weight.

 

Last time I was going out of a day's shooting I dumped my bag into the bathroom scales and tound that it weighed 4 kg - well over target. The extra weight had come from items such as a TURMON monocular, cellphone, pen, and a paperback to read on the train...

 

The open-topped back pocket on my old Domke F6 bag was a great junk swallower, and my wife was forever asking "Is there room in your bag for this?... and this?" I hate to think what the final loaded weight of the bag was.

 

 

 

Best regards,

 

Doug

[pklein]

I attended a wedding reception last weekend. The bride is the daughter of close friends. I've known her for over 20 years, since she was 9 years old. She asked me to bring my camera, so the M8 came along. They had also hired a pro photographer, who uses heavy artillery: A D700 with Sigma 24-70 f/2.8 zoom. I took several lenses, but ended up only using my 35 'Lux ASPH v.1. So the pro had 5 pounds (2.25 kg). around her neck, I had about 2 pounds (950 g).

 

As it got dark, I was shooting at ISO 640 (really 800) at f/1.4, going down to 1/15 to 1/10 second. The pro was shooting at ISO 3200 and f/2.8. So really we were about even (in terms of shutter speed), except that she had more DOF. Eventually we both had to violate our religious principles and use flash.

 

I'm really interested to see the differences in our pictures. The pro avoids shooting wider than f/2.8 because she doesn't want pictures where one eye is in focus and the other not. On the other hand, she has no hesitation shooting at 1600 in any conditions, and 3200 when it's dark. But even with the D700, she seems to prefer keeping things at 3200 and lower. The opposite of many Leica folk, who will shoot wide open anytime, but try to keep the ISO down to 640 (M8) or 1200-1600 (M9).

 

In looking at the pro's posted work, I found that she sometimes shoots at 1600 in the shade, which gives her the ability to stop down and still use motion-freezing shutter speeds.

 

If I had a digital M with a good ISO 3200 and a usable-with-care ISO 6400, I'd be very happy. Can you imagine an M with the D700 sensor, or the new D7000/K5 sensor enlarged to full frame?

 

--Peter

Edited September 9, 2011 by pklein

[roydonian]

My D3 "loses" it between one and two stops higher than my M9. Not 4 stops, not 3 or more. About 1.5...

 

That's interesting. When LFI tested the M9 against the Canon 5D II, its report in issue 8/2009 concluded that the Canon had a two-stop advantage over the M9.

 

The Leica M9 set to 1250 ISO displayed the similar noise level to the Canon set to 5000 ISO, it concluded, but noted “The M9 produces slightly sharper details where the 5D begins to flatten the detail and colour.”

 

(It dismissed the Canon's 12500 and 25000 ISO settings as suffering from "extreme patchiness and severe loss of information.")

 

The ISO speed settings of the M8 were widely reported to be conservative, with 650 ISO equating to a real-world value of 800 ISO. I’ve seen suggestions that this is no longer the case with the M9, but Jamie tells us that the M9 is 1/3 stop more light sensitive than the Nikon.

 

Best regards,

 

Doug

[adan]

Andy.... I'm not following you here on the comment regarding White Balance affecting noise though...... Unless you are actually referring to underexposure in one channel which you are needing to correct in post? Did you mean colour rendiring according to a preferred profile/output rather than white balance?

 

I am referring to correcting the color balance of the image so that white is white and gray is gray (and not a shade of orange or yellow). To me, that is adjusting the White Balance, regardless of whether it is done in the camera (jpegs) or done in post-processing (DNG). In the camera, there is a WB setting - and in any RAW developing software (LR, Camera RAW, C1) there are also WB controls - labelled as such. Doesn't matter which you use.

 

By definition, White Balancing is the correcting of an underexposure in one of the color channels due to the lighting being short of some visible wavelengths. Usually the blue channel, since most artificial light (except electronic flash) tends to the yellowish or yellow-greenish. Basic physics - you have to heat something to over 5000°K for it to put out enough blue light to get "white" light - which is dangerously hot for most lighting units. So most man-made light sources are cooler - and yellower.

 

This results in a gross underexposure of the pixels covered with blue filters. Effectively, under yellow light, if one is shooting at ISO 1250, one is actually exposing the blue pixels three stops less (ISO 20,000). Correcting the WB to neutral, whether the camera does it or you do it, requires amplifying that blue channel ~3x more than the red/green channels. Which adds a substantial amount of noise to the overall picture - since anything that is not actually pure red or green or yellow will have SOME blue component - and include noisy blue pixels. One-quarter of your data (and one-third of the final picture) was shot at ISO 20,000.

 

Use any digital camera at any ISO in daylight - and the images will be far less noisy than images shot at the same ISO in the same camera, but under light with a discontinuous spectrum and corrected for the color imbalance.

 

So, as I said, "Additionally, unless your test shot was under actual daylight, correcting for white balance from other types of light adds its own noise penalty."

Edited September 10, 2011 by adan

[adan]

That's interesting. When LFI tested the M9 against the Canon 5D II, its report in issue 8/2009 concluded that the Canon had a two-stop advantage over the M9.

 

About right - when I shoot the M9 and 5D2 together, I get slightly higher shutter speeds using a Summilux (f/1.4) on the M9 and a 180 (f/3.4) on the 5D2. 2 stops difference.

 

Please let's not involve the Nikon D3 - UNLESS it is the D3X. The D3/D3s has lower noise, but also significantly lower resolution (13 vs. 18 Mpixels). It's like trying to compare Provia 400 to Velvia - sure the Provia is faster, but it captures less detail.

Edited September 10, 2011 by adan

[rosuna]

I do wonder, in its current dire circumstances, how much Kodak is prepared (or is allowed by its bankers) to spend on R&D on upgrading the M9's sensor. I personally would much prefer that Kodak targeted lower noise than higher pixels.

 

I may be wrong but I think Kodak has not much room for improvements. The CCDs are a relatively simple architecture (compared to a typical CMOS).

 

The S2 sensor seems to have incorporated more complexity, mostly for faster data reading (you have to get out those massive 37 millons of "pixels"). But the improvement in noise compared to the M9 sensor doesn't look substantial (there is a difference in pixel size, though).

 

Improvements may be reached outside the sensor itself (other electronics, processing), but it seems Leica will go for a deeper change and for CMOS based alternatives (for better performance or new features). Well informed people in this forum have pointed to this new orientation, and Leica's representatives are suggesting the same (interviews, reportages).

[jaapv]

I would go further than that. I think in the M9 Leica/ Kodak have reached the end of the line of quality improvement on the classical CCD. I would be very surprised not to see new sensor technology like a backlit CMos on the M 10

[Advertisement]

Hallo,

Look here: Leica M

[eronald]

About right - when I shoot the M9 and 5D2 together, I get slightly higher shutter speeds using a Summilux (f/1.4) on the M9 and a 180 (f/3.4) on the 5D2. 2 stops difference.

 

Please let's not involve the Nikon D3 - UNLESS it is the D3X. The D3/D3s has lower noise, but also significantly lower resolution (13 vs. 18 Mpixels). It's like trying to compare Provia 400 to Velvia - sure the Provia is faster, but it captures less detail.

 

I shoot my D3x at 1600 ISO about 100% of the time. At 1600, the images are VERY good. I am thinking of getting a D3s successor just to have two stops more to play with.

 

If a Leica really did good 1600 I would be using one, but I don't think any other camera can compete presently with either the D3s or D3x in their respective classes. The 5DII is a superb camera, but not in the class of the D3x.

 

 

Edmund